Interstitial Optical Diagnosis and Treatment of Breast Cancer

Abstract

Spectral analysis of white light reflected from tissue is a rapid, non-invasive, diagnostic technique. We collected paired optical and conventional histologic measurements from 290 sites in breast and axillary nodes and looked for spectral features to identify cancer. Spectral analysis used artificial neural networks (ANN), hierarchical cluster analysis (HCA) and model based analysis (MBA). The sensitivity and specificity for detecting cancer in breast tissue or lymph nodes were: Breast ANN RCA MBA Sensitivity 69% 67% 94% Specificity 85% 79% 92% Nodes ANN RCA MBA Sensitivity 58% 91% 57% Specificity 93% 76% 85% Therapy aims for complete ablation of small cancers using MR guided Interstitial Laser Photocoagulation (ILP). We have shown that ILP can ablate small cancers and that contrast enhanced MR can detect untreated areas of cancer as small as 2mm. The number of suitable patients for study is small. The main problems are technical rather than biological. Real time imaging of ILP is possible in high field, closed MR scanners (1.5T), where fiber positioning is difficult, but not in low field interventional scanners (O.2T). ILP to fibroadenomas confirmed that laser necrosed tissue is resorbed and the treated area heals safely . The latter could become a valuable treatment in its own right.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 2000
Accession Number
ADA386830

Entities

People

  • Stephen G. Bown

Organizations

  • University College London

Tags

Communities of Interest

  • Advanced Electronics

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Breast Cancer
  • Carcinoma
  • Computers
  • Health Services
  • Lymph Nodes
  • Measurement
  • Medical Personnel
  • Neoplasms
  • Neural Networks
  • Oncology
  • Optical Fibers
  • Optical Properties
  • Pattern Recognition
  • Scattering
  • Surgery

Readers

  • Medical Imaging.
  • Oncology and Biomarker-Based Cancer Detection.

Technology Areas

  • AI & ML
  • Directed Energy